Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-05-23
2004-11-16
Corrielus, Jean M. (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C707S793000
Reexamination Certificate
active
06820084
ABSTRACT:
BACKGROUND
1. Field of the Invention
The invention relates to information processing and communication systems, and in particular to methods and systems for discovering and pushing network information based upon user activity.
2. Background of the Invention
Push technology has become popular in the public eye as an answer to a quantity-of-information problem. An extended digital world we call cyberspace is composed of information sources including the world-wide web, Usenet news, electronic mail, as well as personal data—such as the contents of a user's hard drive and a personal digital assistant.
Cyberspace contains an incredible wealth of valuable information. However, this information is difficult to obtain, and is almost impossible to synthesize and to process. Yet, information on just about anything is available, and it comes in countless different formats. Generally, finding what a user wants requires an enormous amount of searching. Often this search process is so tedious that much of the available information is never sought or found in the first place.
Push technology attempts to solve this problem by selecting relevant information and bringing it to the user. Current push technology attempts to model a user's interests by having a user check off boxes corresponding to various predefined interests. The push technology then uses the model to locate and deliver information corresponding to the checked off interests.
Under such a model, a world of available information is considered to be constantly changing, while the user's interests are assumed to remain fairly static, as indicated by the user interest check list. The perception is that push technology will eliminate the user's need to actively participate in the search for information that he or she is interested in.
And indeed that's true. But one unintended consequence of these models is that much irrelevant information, matching the somewhat primitive selections offered in the check lists, is returned as well. Thus one major problem remaining with current push technology is the problem of separating the wheat from the chaff-sifting through vast amounts of irrelevant information to find that which is pertinent and valuable to a specific user. A significant part of this problem stems from continued reliance upon a user interest model which remains sparse and static.
In part this overload of irrelevant information is due to the availability of faster processors and higher bandwidth data transfer capabilities which allow information to be delivered in a media-rich, TV-like way. Near instantaneous feedback and check-list mechanisms for specifying content are resulting in a large user base for the current push technology.
Recent commercial software applications such as The Pointcast Network™, Marimba's Castanet Tuner™, Netscape's Netcaster™, and Microsoft's CDF channels, as well as several other research and commercial news clipping and information delivery services have provided software which promises to dynamically deliver intranet and Internet information to user desktops.
Under current models, the responsibility for choosing what information gets delivered stays mostly in the hands of the software and the server from which the information comes. Users choose from a series of ‘channels’ or categories, and receive all of the content associated with that category through the service's proprietary software—in the same way that a TV viewer selects a channel and watches all the content provided by that TV channel. The user's choice is limited to the number of checkboxes or channels the push software provides. If the user wants to choose from a larger variety of more specific topics and interests, the model is no longer useful.
Recent estimates suggest that there will be one billion web pages by the year 2000. Even if only one out of every 1000 of those pages is available on an equivalent content channel, users will be required to sift through one million channels to enjoy ‘passive’ delivery of information.
The models created by the current push technology are quasi-static, and require direct intervention on the part of the user to change the topics of interest. Yet, recent research has shown that a user's interests vary at very short intervals—the changes in a user's interests occurring much more rapidly than changes occurring in the information sources from which the information is obtained.
Also, in current push technology, the output of information is limited to current push technology's proprietary output screens. The information is being delivered in a format determined by the information supplier and not by a user's specific needs. Such prepackaged information is often difficult to modify for specific user needs.
A need exists for a push technology that solves the glut of information problem by retrieving information likely to be of interest to a user. A need also exists for a push technology which can provide the information in a form easily adapted to a user's specific format needs.
SUMMARY OF THE INVENTION
These needs—(1) locating information relevant to a user's current interests, and (2) presenting the information in a manner suited to suit a user's specific needs—are met by the present invention, a user-centered push technology.
The invention meets these two critical needs by modeling the user's interests in a dynamic way, and by using the model to locate and to deliver information that remains relevant as the user's needs change.
The present invention is based upon the idea that people and their interests are dynamic, and therefore that the information which a user deems relevant constantly changes because his interests change. Therefore, building and maintaining a reliable, useful user model requires data from numerous sources. What the user is currently browsing, for instance, or topics which he has recently written about or read (in a document, an e-mail, or a new posting), applications he uses, as well as keyboard and mouse activity are all valuable sources of user model information.
In a specific embodiment, the invention defines a method for pushing information to a client in an information processing system. The method includes gathering facts concerning user activity, and using the facts to build a dynamic user model. The method also includes deriving new facts from the facts within the dynamic model, and incorporating the new facts into the model thereby developing probable user interests. Using those interests to locate information either inside the system or outside, such as via a network connection with remote databases. And finally, by pushing the located information to the user in a minimally distracting manner. Alternatively, the information is locally formatted and archived for later use.
Another specific embodiment of the invention defines a system for determining what information is relevant to a user's changing interests, locating such information, and pushing the information to a client triggered on the changes in user interests. The system builds a dynamic user model that contains linked facts defining probable user interests. Independent fact deriving agents elicit new facts from the existing facts, evolving the probable user interests to a level of confidence sufficient for some facts to be pushed to the client. Typically, the pushed facts are displayed for the user, or, alternatively, they are reformatted and archived for later use.
The advantages of a user-centered push technology are that information is located and pushed based on changing user interests, and the information is easily reformatted and archived for later use.
REFERENCES:
patent: 5471629 (1995-11-01), Risch
patent: 5592664 (1997-01-01), Starkey
patent: 5918014 (1999-06-01), Robinson
patent: 6014638 (2000-01-01), Burge et al.
patent: 6134548 (2000-10-01), Gottsman et al.
patent: 6202062 (2001-03-01), Cameron et al.
Barrett Robert Carl
Maglio Paul Philip
Underwood, IV George Milton
Buckley Robert
Corrielus Jean M.
International Business Machines - Corporation
Johnson Daniel E.
LandOfFree
Method and system for pushing information to a client an... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for pushing information to a client an..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for pushing information to a client an... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3286768